N'-(3'-Monophospho-deoxyguanosin-8-yl)-N-acetylbenzidine (dGp-ABZ) is thought to play an important role in initiation of benzidine-induced bladder cancer in humans. This report assesses the possible formation of this adduct by peroxidatic activation of N-acetylbenzidine (ABZ). Adduct formation was measured by 32P-postlabeling. Ram seminal vesicle microsomes were used as a source of prostaglandin H synthase (PHS). The peroxidatic activity of PHS was compared to that for horseradish peroxidase. Both peroxidases converted ABZ to dGp-ABZ whether DNA or 2'-deoxyguanosine 3'-monophosphate (dGp) was present. Following 32P-postlabeling, the enzymatic and synthetic adduct were extracted from PEI-cellulose plates and shown to have the same HPLC elution profiles for the bisphosphate adduct (32P-dpGp-ABZ). Treatment of the enzymatic and synthetic bisphosphate adduct with nuclease P1 yielded a product which eluted at the same time from the HPLC (32P-dpG-ABZ). Additional experiment s demonstrated that the PHS-derived 5'-monophosphate (dpG-ABZ ) and 3'-monophosphate (dGp-ABZ) adducts were also identical to their corresponding synthetic standard. With comparable amounts of total ABZ metabolism, PHS produced about 40-fold more dGp-ABZ than horseradish peroxidase (1943 + 339 vs 49 + 7.8 fmol/mg dGp). Adduct formation was dependent upon the presence of peroxidase and the specific substrate, i.e., arachidonic acid or H2O2. Adduct formation by PHS was inhibited by indomethacin (0.1 mM), ascorbic acid (1 mM) and glutathione (10 mM), but not by DMPO (100 mM), a radical scavenger. Horseradish peroxidase adduct formation was also inhibited by ascorbic acid and glutathione. In addition, DMPO elicited greater than a 96% inhibition. Results demonstrate peroxidatic metabolism of ABZ to form dGp-ABZ. The mechanism of dGp-ABZ formation by PHS and horseradish peroxidase may be different.